Sridhar wrote...
> S/390 does NOT run on microprocessors of any kind, let alone
> one as slow as PPC.
Don't tell that to the folks at http://www.conmicro.cx/hercules/
(Of course, it depends on what your definition of "on" is).
Brian
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
_| _| _| Brian Knittel / Quarterbyte Systems, Inc.
_| _| _| Tel: 1-510-559-7930 Fax: 1-510-525-6889
_| _| _| Email: brian(a)quarterbyte.com
_| _| _| http://www.quarterbyte.com
On December 16, Innfogra(a)aol.com wrote:
> Found it. A very interesting issue. There are several other stories I would
> be interested in also. Thank you.
You're very welcome.
Unfortunately these scans aren't from *my* bytes...my collection
doesn't go back that far, as I was 7 years old in 1976...so I can't go
scan more of 'em. :-(
Eventually I'll beef up my BYTE collection, probably via eBay. There
are a LOT of great articles in there. Of course I do have all of the
wonderful Ciarcia books, so my BYTE collecting motivation has been
partially satisfied. :-)
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On December 16, ajp166 wrote:
> > I wonder if it would be possible (and practical) to use a
> >microcontroller, perhaps a PIC, to act as a core controller. Use the
> >A/D and D/A hardware to handle the drive and sense stuff, and do all
> >the timing in firmware...making it easily tweakable.
>
> Not likely as the sense voltages are quite small and the slice time
> has to just right.
Well I wouldn't expect to connect it *directly* to the PIC...some
analog jellybeans would seem appropriate.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On December 16, Ben Franchuk wrote:
> > > Remember the (iirc) Compaq ad. Thier 286 had a meg of static ram onboard.
> > > The ad depicted an empty desk with two tire tracks burt into it and a
> > > surprised user behind.
> >
> > Static RAM? Are you sure? I've never heard of static RAM in a
> > PeeCee. That's neat.
>
> Yes there is some -- it is called cache.:) I suspect a sever type
> PC is more likely to have static than dynamic. I never did like PC
> marketing -- clock speed -- not memory speed used as a benchmark.
Nonono, I mean *main* memory, not cache. I sincerely doubt there
are any PeeCees in the 486-or-newer arena with static RAM for main
memory.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
>> I have some scans of an old BYTE Magazine article (July 1976 I
think)
>> that details how to build a common interface for an arbitrary chunk of
>> core. I haven't really read it in-depth but I plan to someday. I can
>> put it online if anyone would like to see it.
I'd love to see scans of this one too. I remember the article but
somehow lost it.
Allison
From: Don Maslin <donm(a)cts.com>
>It should not have a twist like a PC cable, and can be a length of
>34-conductor ribbon cable, Gene, with a 50-pin female header on each
>end. What ever length is needed.
Ah, dont you mean 34 pin? 50 pin would be the SCSI.
Allison
The older large ferrite core is easier to work with though
much slower. The bigger cores produce a larger output
when they switch but the cycle times are in the
3-5uS range. The later is helpful for demos as nothing
is too fast.
Allison
-----Original Message-----
From: Loboyko Steve <sloboyko(a)yahoo.com>
To: classiccmp(a)classiccmp.org <classiccmp(a)classiccmp.org>
Date: Saturday, December 15, 2001 11:36 PM
Subject: Re: [PDP8-Lovers] how to clean a PDP8/A, dishwasher?
>I've looked into this too. Problems are many. It's
>definitely not a trivial project, lots of analog and
>electromagnetic voodoo, and the chips that made it
>easier are very hard to find. I've got a small
>capacity large doughnut ca. 1960 core plane from an
>IBM machine and a 4K by 16 plane from the 70's with
>very tiny doughnuts, both unused, and I'd love to
>demonstrate how it worked.
>
>
>--- ajp166 <ajp166(a)bellatlantic.net> wrote:
>> From: John Allain <allain(a)panix.com>
>> >I have a Question for the other core users out
>> there:
>> >How would I test-signal a board, for demos, that is
>> >just a core frame, IE one sacrificed from its stack
>>
>> >and sold at the e- flea market?
>>
>>
>> It's a non trivial thing to do. Core by definition
>> is destructive
>> read out memory. So to demo a core you need to
>> provide
>> the coincident current (x,y) and the
>> inhibit/write/read signals
>> with the associated timing. Both the currents and
>> the
>> timing are critical. takes a lot of stuff to do
>> that.
>>
>> Allison
>>
>>
>
>
>__________________________________________________
>Do You Yahoo!?
>Check out Yahoo! Shopping and Yahoo! Auctions for all of
>your unique holiday gifts! Buy at http://shopping.yahoo.com
>or bid at http://auctions.yahoo.com
From: Gene Buckle <geneb(a)deltasoft.com>
>I know I can use any length of ribbon I want. What I'd like to know is
>what it was when it was new. (and it's 34 pin card-edge connectors that
>are needed :) )
I take them off old PC floppy cables.
>The drive is a Miniscribe 8425S and AFAIK, it's set to ID 0 with the
>terminating resistors installed. Seagate or Maxtor is not listed as a
>_controller_ choice when HFORMAT is run.
Yep! It's possible to use that drive but the Ampro provided tools only
recognized drives from the era, there were many good choices in
later years.
Allison
On December 15, Ben Franchuk wrote:
> > I have some scans of an old BYTE Magazine article (July 1976 I think)
> > that details how to build a common interface for an arbitrary chunk of
> > core. I haven't really read it in-depth but I plan to someday. I can
> > put it online if anyone would like to see it.
>
> I think that was a design for the 8080 and the S100 bus.
> It was a really nice design too if I remember right. I remember
> somebody doing a core-demo on the web. I think it used a iron washer
> as the core. Since this is not high speed only the current is critical
> here.
It should be pretty easy to interface to other processors, then...
> (Boy do I miss the early Bytes -- ordinary people designing and using
> computers
> -- not multi-billion $$$ bloatware companies -- )
Yup.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
> -----Original Message-----
> From: Jochen Kunz [mailto:jkunz@unixag-kl.fh-kl.de]
> On 2001.12.14 18:08 Christopher Smith wrote:
>
> [Sequent Symmetry S81]
> > Drives, but I can probably handle that...
> What interface SMD?
Possibly, it's hard to tell since I'm not sure what an SMD interface looks
like. I could turn it over and check but I don't know what to look for. ;)
> > I believe that the system was board-for-board compatible
> with the S27,
> The Unix-AG once owned a S27 with 8 CPUs, 80MB RAM, 7 SMD disks,
> Exabyte, ... I know where the machine is now, including a complete set
> of OS software. If you get the machine working, ask me for software.
I certainly will.
> Don't expect much of that machine. I could get only 200kB/s via the
> Ethernet, the SMD disk controller could not handle more then 700kB/s.
> (The same disks can deliver more than 2MB/s on a Sun 3/260.) That is
> real poor for such a high end beast. I was really
> disapointed. But it is
Did you get the chance to find out how performance faired when you add more
CPU?
> nice to see an old 4.2BSD UNIX derivate with AT&T universe
> running on 8
> CPUs in a SMP config.
It is amusing for me to see _any_ intel cpu running in an SMP config, and
the fact that from what I've heard, Sequent made them do it fairly well,
intrigues me.
Regards,
Chris
Christopher Smith, Perl Developer
Amdocs - Champaign, IL
/usr/bin/perl -e '
print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");
'
In a message dated 12/15/01 10:54:50 PM Pacific Standard Time,
mcguire(a)neurotica.com writes:
>
>
> The scans are up at http://ti.neurotica.com/core. They're a bunch
> of jpegs. No, that wasn't my idea, I didn't do the scanning. :) I
> might try to turn them into a pdf one of these days.
>
>
Found it. A very interesting issue. There are several other stories I would
be interested in also. Thank you.
Paxton
Astoria, OR
Anyone have any of the docs for this beastie? I have one but no docs. I
have been having trouble figuring out which terminal emulation to use when
connected to it, also If someone has the docs I would be will ing to get my
scanner hooked up again to my computer and see about making a PDF of them
for posterity....
In a message dated 12/15/01 9:15:47 PM Pacific Standard Time,
mcguire(a)neurotica.com writes:
> I have some scans of an old BYTE Magazine article (July 1976 I think)
> that details how to build a common interface for an arbitrary chunk of
> core. I haven't really read it in-depth but I plan to someday. I can
> put it online if anyone would like to see it.
>
I would like to see it. I have several pieces of core.
Paxton
Astoria, OR
On December 16, Pat Finnegan wrote:
> > I have some scans of an old BYTE Magazine article (July 1976 I think)
> > that details how to build a common interface for an arbitrary chunk of
> > core. I haven't really read it in-depth but I plan to someday. I can
> > put it online if anyone would like to see it.
>
> I'd definately be interested.
The scans are up at http://ti.neurotica.com/core. They're a bunch
of jpegs. No, that wasn't my idea, I didn't do the scanning. :) I
might try to turn them into a pdf one of these days.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On December 16, John Allain wrote:
> > I have some scans of an old BYTE Magazine article (July 1976 I think)
> > that details how to build a common interface for an arbitrary chunk of
> > core. I haven't really read it in-depth but I plan to someday. I can
> > put it online if anyone would like to see it.
>
> It Would help lots more Dave if you dug out that disk drive.
> The mv3500 is still pretty useless w/o swap space.
I've gotten to the bottom of the "staging area"...I think that pile
of drives is in my storage locker up in Maryland. I will know for
sure soon.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On December 15, ajp166 wrote:
> >I have a Question for the other core users out there:
> >How would I test-signal a board, for demos, that is
> >just a core frame, IE one sacrificed from its stack
> >and sold at the e- flea market?
>
> It's a non trivial thing to do. Core by definition is destructive
> read out memory. So to demo a core you need to provide
> the coincident current (x,y) and the inhibit/write/read signals
> with the associated timing. Both the currents and the
> timing are critical. takes a lot of stuff to do that.
I have some scans of an old BYTE Magazine article (July 1976 I think)
that details how to build a common interface for an arbitrary chunk of
core. I haven't really read it in-depth but I plan to someday. I can
put it online if anyone would like to see it.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On December 15, Ben Franchuk wrote:
> > > > I have little doubt about the appropriateness... of your response.
> > > > Care to say when its appropriate to own three S/390's?
> >
> > > It's wholly inappropriate. He should give one to me.
> >
> > And one to me. Then he has one, and that's somewhat less inappropriate. ;-)
>
> Why collect 390's? 360's are more impressive! :)
I suspect Sridhar doesn't have the S/390s for "historic"
purposes...Ordinarily I'd just shut up and let him say that, but I
just talked with him on the phone, he's outside of Richmond on his way
down here. I'm going to make him eat some beef while he's here. :-)
> BTW does the 390 architecture still emulate the 360's?
I believe so, yes.
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
From: Richard A. Cini, Jr. <rcini(a)msn.com>
> On the negative side, a small regression error has crept in and 4k
BASIC
>will no longer run. This is not so bad because if you could run 8k BASIC
>there's no reason to run 4k BASIC.
This is broken. Since 4kbasic and 8kbasic ran on the same altairs it
suggests something in the emulator is broken.
FYI: much of MITS and Altair compatable software used input Port 0FFh
as a 8bit input (usually to configure IO).
Allison
I only needed one floppy and with the twisted end cut off
it was perfect.
Allison
-----Original Message-----
From: Richard Erlacher <edick(a)idcomm.com>
To: classiccmp(a)classiccmp.org <classiccmp(a)classiccmp.org>
Date: Saturday, December 15, 2001 11:00 PM
Subject: Re: Ampro Little Board questions...
>A PC floppy cable certainly wouldn't work with the Little Boards I have,
since
>they use the standard floppy disk cable, which has no twists or cuts.
>
>Dick
>
>----- Original Message -----
>From: "ajp166" <ajp166(a)bellatlantic.net>
>To: <classiccmp(a)classiccmp.org>
>Sent: Saturday, December 15, 2001 2:44 PM
>Subject: Re: Ampro Little Board questions...
>
>
>> From: Gene Buckle <geneb(a)deltasoft.com>
>> >belongs with it. Does anyone have a Series 100 box that could give
me
>> >the measurements and connector orientation & location of the floppy
>> >cable?
>>
>>
>> Nothing magic. the last one I had used was from a PC.
>>
>> >Secondly, I don't know what "brand" SCSI controller is built into the
>> >board. The SCSI chip seems to be an NCR 5830, but that leads me
nowhere
>> >in relation to whether or not it's an Adaptec or other model
controller.
>> >I need to know the brand because the hd formatting software needs to
>> >know it.
>>
>>
>> Your applying PC logic to it. It's is not Adaptec, i'ts just SCSI
>> (SCS1 or II) host and the NCR5380 is one of the early and common
>> chips used for that.
>>
>> The brand applies to the "other" board, what is known as a SCSI
>> bridge board. Adaptec, Xybec, WD and other made them. You
>> need to know what board and what drive to do the formatting. If
>> you dont mind hacking Z80 code you can go frm an AMPRO LB
>> with SCSI to a smaller SCSI drive (64mb or less, or the rest will
>> be unused).
>>
>> Allison
>>
>>
>>
>>
>
>
From: John Allain <allain(a)panix.com>
>I have a Question for the other core users out there:
>How would I test-signal a board, for demos, that is
>just a core frame, IE one sacrificed from its stack
>and sold at the e- flea market?
It's a non trivial thing to do. Core by definition is destructive
read out memory. So to demo a core you need to provide
the coincident current (x,y) and the inhibit/write/read signals
with the associated timing. Both the currents and the
timing are critical. takes a lot of stuff to do that.
Allison
Hello, all:
Just a quick announcement. Today was the official release of version 2.1 of
the Altair32 Emulator. Yeah!
It's been about four months since the 2.0 release and a lot has happened
since then. Floppy disk images for CP/M, Altair DOS and Disk BASIC are fully
functional and the Altair32 now has an integrated debugger, courtesy of Jim
Battle's Sol-20 emulation project. These two items alone took over half of
the time between releases and integrating the debugger required a major
overhaul of the 8080 emulation code.
On the negative side, a small regression error has crept in and 4k BASIC
will no longer run. This is not so bad because if you could run 8k BASIC
there's no reason to run 4k BASIC.
The next phase of the project will hopefully include some or all of the
following:
Continued code cleanup; minor fixes to IMSAI conditional
Further testing of the debugger (it's been only lightly tested with the
Altair32)
A set of Windows-based tools to manipulate diskette images (to enable people
to download programs from http://www.retroarchive.org and get them onto
diskettes)
Support for BBS software
Support for different floppy controllers/disk formats and maybe a
pseudo-floppy hard drive.
That's it for now. As always, you can check out the project at:
http://highgate.comm.sfu.ca/~rcini/classiccmp/Altair32.htm
Rich
Rich Cini
Collector of classic computers
Build Master for the Altair32 Emulation Project
Web site: http://highgate.comm.sfu.ca/~rcini/classiccmp/
/************************************************************/
On Dec 15, 12:28, Matthew Sell wrote:
> Typically, the time spent "in the water" isn't long enough to damage.
Even
> items made of steel and iron won't rust if the water is removed after the
> cycle is complete. If they sit overnight, well, that's a different story.
>
> The only production problem I saw with untreated water was with an
> electronic test instrument that had a lot of high-impedance signal
> interconnects throughout. Many signal lines ran for long distances next
to
> each other. While the design of this piece of test equipment was
> questionable, it was our duty to get it to work.
>
> The two biggest problems were contaminants from the water supply used in
> the washing process (city water - switched to using a commercial
filtration
> system), and humidity (had to paint a sealant on all of the boards).
That's a well-known problem. Some of the residues from a domestic water
supply -- especially in hard water areas -- are mildly hygroscopic, and as
a result, the boards would acquire very small amounts of moisture on the
surface, especially when exposed to a humid atmosphere. In combination
with the salts in the residue, this makes for leakage across the board,
which could easily upset very high impedance circuits.
I heard of someone who had the opposite problem. He designed a CMOS
circuit which worked fine when forst contructed, but stopped when given
anti-environment protection or was potted. He'd inadvertantly relied on
the normal leakage across a PCB to hold the unused inputs of a CMOS gate at
a particular level. Remove the leakage current and the gate stops working
properly. Solution: add the pullup resistor that should have been there in
the first place.
--
Pete Peter Turnbull
Network Manager
University of York
On Dec 15, 1:32, Matthew Sell wrote:
> Here's the secrets to doing this right:
>
> 1) Do not use any detergents. NONE.
Why not? Commercially, detergents are used to remove flux. I use
laboratory detergent to clean PCBs I've made or modified. I don't use
washing-up liquid, though, and I don't use the harsher detergents sometimes
found in dishwasher detergent.
--
Pete Peter Turnbull
Network Manager
University of York
On December 15, Jim Davis wrote:
> IMHO: All software development should be performed as such:
>
> 1) Requirements - what should it do, and not do. Spin this till
> everybody
> signs off.
> 2) Prelim design - Ok, a rough outline of the design, data structures
> and control/data flow defined here.
> 3) Detailed design - Define all the modules and their function, break it
> down.
> 4) Test plan - integrate testing into detailed design, make it unit
> testable.
> a unit is somthing that has input and output and side effects, like a
> function.
> 5) Finally, coding - build modules in parallel with test code.
> 6) Unit testing - verify that modules comply with detailed design.
> 7) Integration testing - hook it all together, make sure it works, apply
> test plan developed in step 4 for fully integrated aplication.
>
> Do 1-3 until marketing decides what they want,4-7 until you find no
> errors.
>
> For safety critical, you should /have to perform statement and decision
> coverage in
> step 6 and 7 and the detailed design should have a one-to-one
> corespondence
> with the detailed design document.
Hmm, that procedure "reads" nice, but it sounds like more meetings
than actual work. But then I've been a software developer for about
twelve years, and nobody that I've worked with can figure out how I
can blow off all the meetings and not get fired...it's because I end
up writing all the code that the rest of the developers are talking
about in their meetings, WHILE they're in their meetings, and by the
time they're doing screwing around, the code is running.
Procedures are nice, but they can be taken too far. Goal-orientation
is better.
(While I'll freely admit that this approach simply doesn't work for
multi-million line applications, I should state that I generally work
on applications of less than one million lines, but generally more
than 100,000. I further state that my methods should probably NOT be
used in life-critical applications...more than two eyes need to look
at that stuff, no matter what.)
-Dave
--
Dave McGuire
St. Petersburg, FL "Less talk. More synthohol." --Lt. Worf
On Sat 15 Dec 2001, Jeffrey S. Sharp wrote:
> There's a big difference between writing code to solve problems and being
> a software engineer. Designing, coding, and compiling is only 40% of the
> battle. Hopefully you're also spending some time planning and testing.
The company I work for seem to have forgotten the planning and testing part
of the software design process. We had a lot of embedded software written by
outside contractors for a 486 running QNX real time OS. Over the years the
software has evolved into the hardware equivalent of a rats nest and it's been
left to our customers to find the bugs - most being "show stoppers".
I sometimes wonder just how many customers we've lost because of this.
Also, dont get me wrong, the same should be applied to hardware design.
We recently interviewed an electronics engineering graduate who didn't know
the difference between NPN and PNP transistors !! What do they teach kids
these days ??
Should engineers be licensed ? - It's not a bad idea.
Chris Leyson
